迷失在潛在空間：關於物理模擬的潛在擴散模型實證研究

摘要

擴散模型在推理階段的高昂計算成本阻礙了其作為快速物理模擬器的應用。在圖像和視頻生成的背景下，這一計算缺陷已通過在自動編碼器的潛在空間而非像素空間中生成內容得到解決。在本研究中，我們探討了類似策略是否能夠有效應用於動力系統的模擬，以及其代價幾何。我們發現，潛在空間模擬的精度對於廣泛的壓縮率（高達1000倍）表現出驚人的穩健性。我們還展示了基於擴散的模擬器在精度上始終優於非生成式模型，並通過更高的多樣性來補償其預測中的不確定性。最後，我們討論了從架構到優化器等實際設計選擇，這些選擇對於訓練潛在空間模擬器至關重要。

English

The steep computational cost of diffusion models at inference hinders their use as fast physics emulators. In the context of image and video generation, this computational drawback has been addressed by generating in the latent space of an autoencoder instead of the pixel space. In this work, we investigate whether a similar strategy can be effectively applied to the emulation of dynamical systems and at what cost. We find that the accuracy of latent-space emulation is surprisingly robust to a wide range of compression rates (up to 1000x). We also show that diffusion-based emulators are consistently more accurate than non-generative counterparts and compensate for uncertainty in their predictions with greater diversity. Finally, we cover practical design choices, spanning from architectures to optimizers, that we found critical to train latent-space emulators.

迷失在潛在空間：關於物理模擬的潛在擴散模型實證研究

Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulation

摘要

Support